AI-Token Rally: NEAR, FET and Bittensor Draw Fresh Flows

Crypto markets saw a fresh burst of interest in decentralized‑AI tokens between May 21 and May 22, 2026, as NEAR, Fetch.ai derivatives (FET) and Bittensor (TAO) posted notable intraday gains and volume spikes after a string of announcements and renewed investor attention to on‑chain AI markets.

NEAR Protocol led many headlines on May 22, jumping roughly 20–26% in a single day as traders pinned gains to the token’s positioning at the intersection of identity, privacy and AI infrastructure. Several market trackers and news outlets flagged a sudden rotation into AI‑adjacent infrastructure names that day.

Fetch.ai (FET) and the wider Artificial Superintelligence Alliance grouping also logged renewed buying pressure, with FET trading higher on May 22 amid coverage of new product features and platform updates that analysts said could broaden on‑chain AI utility. Crypto pricing feeds and market updates recorded elevated volume and intra‑day gains for FET during the same window.

Bittensor’s TAO — a token tied to a permissionless, incentive‑driven decentralized ML network — showed active flows as well, with price and volume data displaying volatility across May 21–22 as traders reacted to subnet growth, governance chatter and protocol updates. Native outlets and historical price feeds logged the swings and heavier trading activity.

Market commentators framed the moves as a resurgence of speculative capital into the decentralized‑AI theme: investors are testing whether decentralized compute, data‑market primitives and on‑chain inference layers can capture parts of the broader AI supply chain being built by cloud and silicon incumbents. Several daily market roundups explicitly tied the rallies to that narrative.

Traders and journalists pointed to a handful of catalysts that helped concentrate flows. Those included fresh product updates from AI‑focused projects, positive industry headlines (including outsized tech earnings that rekindled AI appetite), and on‑chain developments such as subnet launches or governance proposals. Coverage suggested the moves were more narrative‑ and flow‑driven than the result of a single protocol upgrade.

Technical and on‑chain mechanics specific to these projects shaped how the news translated into prices. Bittensor’s subnet model and dynamic TAO mechanisms, plus staking and emissions debates, created paths for both speculative entry and concentrated selling by major holders. Analysts warned these design choices can amplify volatility in tokens that are still finding product‑market fit.

Volume and order‑book dynamics amplified short‑term moves. Coin and exchange feeds showed outsized 24‑hour volumes for NEAR and FET on May 22, and reports flagged whale buying, derivatives open interest and short‑covering as contributors to the spikes. Those same dynamics explain why intraday gains often retrace quickly in this sector.

The episode exposed familiar tradeoffs for decentralized‑AI proponents: tokenized incentives can accelerate network growth and align contributors, but they also make protocols vulnerable to narrative swings and governance‑linked selloffs. Recent reporting singled out specific governance episodes and large holders’ activity as near‑term risks for Bittensor in particular.

Beyond price action, some researchers and builders argue decentralized markets could carve durable niches in data exchange, model marketplaces and inference coordination — especially where privacy, data provenance and composability matter. Academic and practitioner work on on‑chain benchmarks and DAO‑governed AI systems points to both technical promise and governance complexity.

For traders and institutions the May 21–22 moves were a reminder that the decentralized‑AI narrative remains potent but unevenly realized: markets are now testing which protocols deliver measurable on‑chain revenue and which remain story‑led. Observers said the immediate takeaway was that speculative flows can re‑ignite quickly when the broader tech narrative turns back toward AI, but that long‑term value will require demonstrable product usage and clearer token economics.